Nvidia and AMD seek Trump help on China chip controls | Magica
Nvidia and AMD seek Trump help against congressional chip controls
Editorial Team
••📖8 min read
Nvidia and AMD have sought White House help against bipartisan chip-export restrictions in the defense bill, turning access to China into a contest over congressional oversight and presidential discretion.
Nvidia and AMD have sought administration intervention against China chip restrictions in the annual defense policy bill.
Their campaign distinguishes between opposing tighter congressional oversight and negotiating the mechanics of chip tracking.
Trump’s support offers leverage, but restrictions have bipartisan backing and supporters inside his administration.
U.S. AI-chip suppliers Nvidia and Advanced Micro Devices have sought President Donald Trump’s help to stop Congress from tightening controls on exports to China. Their campaign puts presidential discretion over chip sales against bipartisan demands for stronger statutory limits and oversight.
The immediate battleground is the National Defense Authorization Act, the annual defense policy bill. But the administration already imposes licensing safeguards, and the Senate proposal’s sponsor presents it as reinforcing Trump’s policy.
A presidential route around congressional resistance
The companies increased outreach to the White House and senior Commerce Department officials after months of discussions on Capitol Hill, according to a September 29 report citing three people familiar with their strategy. Their request was for the administration to pressure lawmakers against including China-related export restrictions in that bill. One person described the lobbying as still in its early stages, initially addressing technical language with lower-level officials.
Jensen Huang, Nvidia’s founder and chief executive, has argued that excluding American suppliers from China would encourage Chinese chip development, sacrifice a large market and weaken the U.S. position in AI, according to the September report. His company supplies the computing hardware used to build AI systems; its commercial exposure gives that strategic argument a direct financial stake.
The relationship has already produced a precedent for seeking presidential intervention. The September report said Trump overruled opponents in Congress and his own administration late in 2025 to permit Nvidia and AMD to sell some advanced chips to China in exchange for a share of profits. It also described Huang putting Trump on speakerphone at a technology conference in September, where the president voiced support.
The Cambridge startup has secured backing described as eight figures in capital and compute, including UK government support. Its wager is that cheaper, accurate quantum calculations can supply better AI training data; full performance results remain forthcoming.
Editorial Team
But access has not settled the underlying dispute. Lawmakers seeking tighter restrictions say supplying powerful chips helps Beijing compete in AI and increases security risks. Senator Elizabeth Warren, a Massachusetts Democrat who has pressed for restrictions in the defense bill, has challenged both the sales and Huang’s influence with Trump.
A November 3 analysis identified Huang’s relationship with the president and efforts to preserve China sales as a source of exposure to renewed congressional scrutiny. It also stressed that Trump’s continued executive authority could shield business allies even under a Democratic Congress. The assessment was conditional; it did not establish an election result or a subsequent change in export policy.
The existing policy supplies an important baseline. In a January 13, 2026 announcement, Commerce’s Bureau of Industry and Security said applications to export Nvidia H200, AMD MI325X and similar chips to China would receive case-by-case review. Applicants must demonstrate that exports will not reduce global semiconductor production capacity available to U.S. customers, that Chinese purchasers have export-compliance procedures, and that products undergo independent testing in the United States.
Their principal target, according to the September report, was language drawn from the AI OVERWATCH Act, introduced by House Foreign Affairs Committee Chair Brian Mast, a Florida Republican, and Senator Jim Banks, an Indiana Republican. The House and Senate versions differ, but industry representatives objected to both versions’ tighter oversight of chip sales to China and other adversaries.
Banks’ own July 14 description complicates any clean division between Trump and Congress. He presented the Senate measure as codifying Trump’s restrictions, including an 18-month extension of a prohibition on the most advanced chips going to countries of concern. For less advanced chips, his office described a requirement for Commerce to certify to Congress that exports would not undermine U.S. supply, divert global foundry capacity, breach end-user or end-use controls, or enable unauthorized remote access.
The Senate proposal would also require licenses for exports, reexports and transfers of advanced chips to countries of concern, prohibiting blanket approvals, and direct Commerce to develop a government strategy for competing with China in AI. These are the sponsor’s descriptions of the Senate provisions, rather than a complete account of both chambers’ texts.
Banks said his earlier GAIN AI Act passed the Senate the previous year and its principles—prioritizing American customers for advanced chips—were incorporated into the administration’s January H200 rule.
The disagreement therefore reaches beyond whether some exports should be allowed. It concerns how much discretion remains with the administration, and what assurances Congress can demand before licenses are approved.
On the Chip Security Act, the companies saw more room for negotiation. Industry officials wanted clarity that tracking requirements would not mandate hardware changes, a clearer assignment of responsibility within the supply chain, and more time to comply, the September report said. Those requests bear on the practical burden of tracing chips after sale, rather than simply on permission to export them.
A lobbying disclosure lists the Chip Security Act, AI OVERWATCH Act, Remote Access Security Act and fiscal 2027 defense authorization among its trade-policy issues. It identifies Congress, Commerce, the National Security Council and White House economic and science offices among the institutions contacted on that issue area. The record documents activity across institutions; it does not establish support from them or identify spending attributable to this particular campaign.
Company-reported figures from Nvidia show its H20 inventory charge alongside overall revenue growth in the quarter ended April 27, 2025. Source: Nvidia.
A financial cost and a competing security case
Export policy has already imposed measurable costs on Nvidia. In its May 2025 earnings release, the company said new U.S. licensing requirements for H20 products destined for China led to a $4.5 billion charge for excess inventory and purchase obligations in the quarter ended April 27, 2025.
H20 products were designed primarily for China. Nvidia reported $4.6 billion in H20 sales before the requirements—about 10% of its $44.1 billion total quarterly revenue, calculated from the reported figures—and said it could not ship an additional $2.5 billion in H20 revenue that quarter. The inventory charge and unshipped revenue measure different things and should not be added into a single loss figure. They nevertheless show why control over licensing is a material business issue, rather than a debate confined to hypothetical future demand.
The company also forecast approximately $8 billion in lost H20 revenue from export limitations in the following quarter. That was a May 2025 forecast, not an observed result or a measure of the pending legislation’s cost. Even with restrictions, Nvidia reported total quarterly revenue up 69% from a year earlier: substantial China exposure coexisted with rapid overall growth.
AMD’s 2025 annual report, filed February 4, 2026, shows why restrictions should be distinguished from permanently lost business. AMD recorded approximately $800 million in inventory and related charges on its Instinct MI308 data-center graphics processors in the second quarter of 2025 because of U.S. export restrictions. After receiving some licenses and resuming shipments to certain Chinese customers in the fourth quarter, it reversed approximately $360 million of those charges.
AMD also said officials expected the U.S. government to receive 15% of revenue from licensed MI308 sales to China, but no regulation establishing that requirement had been published as of the filing. That qualifies the September report’s description of a profit-sharing arrangement: AMD’s account referred to revenue and distinguished officials’ expectations from a published requirement.
The security rationale also predates the present lobbying campaign. In October 2022, Commerce’s Bureau of Industry and Security, the agency administering export controls, announced restrictions on China’s access to advanced computing chips, supercomputers and semiconductor manufacturing capabilities. It cited their use in military systems, military decision-making and surveillance.
Huang’s argument is that restrictions can undermine American influence by accelerating alternatives in China. The opposing argument is that denying access to American hardware can constrain Chinese capabilities. The available evidence does not establish how much either effect outweighs the other. Nvidia’s financial exposure demonstrates its incentive to preserve sales; it does not resolve the national-security calculation.
Company-reported AMD figures show the original MI308 charges and the portion later reversed after licensed shipments resumed. The reversal is part of the earlier charge. Source: AMD annual report.
The defense bill remains the test
The September reporting anticipated House-Senate negotiations after the November election and completion of the defense package by year-end. The substantive test is which provisions survive: an extended restriction on the most advanced chips, certification conditions for other exports, and practical obligations for tracking chips after sale.
Even within the administration, the companies face uneven support. An industry representative described sympathy from some White House and Commerce officials but greater resistance at the Pentagon and the U.S. trade representative’s office. Nvidia, AMD, the White House and Commerce did not respond to requests for comment for that report.
The evidence needed next is a negotiated legislative text and an administration position on it. Presidential access alone does not show whether Congress will impose additional statutory limits, or how Commerce would apply the resulting licensing and tracking requirements.
Japan’s Nikkei fell 7.09% in Friday morning trading, with AI, chip and data center-related shares hit hard after Wall Street losses. Reports linked the rout to fears of U.S. regulation following Democratic election victories, without identifying an enacted restriction or a cut in AI spending.
Potential investors are pressing Anthropic for a lower valuation as a technology selloff threatens its November IPO timetable. Rapid revenue growth and substantial cash complicate a financing debate driven by losses and long-term infrastructure commitments.
Visa's new benchmark finds that AI shopping models often miss customer requirements and struggle to repeat successful decisions. The tests cover product discovery, leaving checkout and payment reliability for future rounds.
Agility Robotics began trading after its Churchill Capital Corp XI merger, bringing a commercially deployed humanoid robot maker to U.S. public markets. Its next test is turning conditional Digit 5 orders into deliveries and sustainable economics.
Goldman Sachs sees temporary halts potentially affecting about 45% of planned US data center expansion after Democratic election victories. In Texas, where approvals were already paused, the next fight is over how long restrictions last and what developers must pay for.
Core Scientific’s third-quarter colocation revenue rose 57.6% to $215.4 million. Its expansion beyond CoreWeave could diversify sales from 2027, but the next projects require additional financing rather than the same customer funding that supported its existing conversions.
Meta is rolling out voice conversations and animated avatars for Muse in the US. The update adds personality to a cloud-based task agent whose costs, permissions and reliability matter more than its expressions.
Anthropic is rolling out Personal Finance to paid Claude users in the US, using Plaid to read bank data. Its pitch rests on ad-free answers and privacy controls, while its own documentation warns that balances and budgets can be incomplete.
OpenAI president Greg Brockman predicts significant medical advances within 12 to 18 months and raises the possibility of death becoming a choice. OpenAI’s cell research and a separate early drug trial show narrower progress, with financial ties, safety questions and clinical testing still central to judging the promise.
Citrini Research is recasting its AI investments around relative value as it warns that financing and construction constraints threaten returns. Its reported Oracle–Nebius trade faces a crucial counterargument: Nebius has substantial customer commitments, although delivery remains conditional.
Shield AI has agreed to acquire Echodyne for up to $5.3 billion, backed by a $2.5 billion financing package. The radar maker brings a profitable, defense-heavy business and commercial customers, with a promise to keep supplying outside system builders.
Gemini Omni 1.2 Flash lowers Google's quoted 720p generation price to $0.07 per second and extends native clips from 10 to 15 seconds. Editing limits and early commercial tests show why cheaper output does not guarantee the same savings on finished videos.
Mercedes-Benz plans to add Liquid AI’s on-device models to its North American voice assistant. The deal combines local processing with existing cloud AI, while customer availability, measured performance and contract economics remain unconfirmed.
Mistral is expanding Studio with managed agents, apps and workflow tools, extending its existing enterprise platform. Its private-deployment pitch sits alongside cloud-hosted app services and vendor-stored agent sessions, making the chosen configuration central to customer control.
Meta president Dina Powell McCormick expects tougher conditions for new US data centers after Democratic election victories, but says states still want investment. The test is whether electricity commitments, local jobs and tax benefits can secure community support.
OpenAI is releasing GPT-6 Cyber to selected verified defenders with stronger monitoring. Its Linux research shows the scale and cost of automated exploitation, while mixed benchmark results and unfinished patching limit what the launch proves about safer software.
Black Hills plans $1.8 billion of gas generation under agreements to serve Google’s proposed $17 billion Cheyenne campus through 2048. Google has advanced equipment funding, but delivery, fuel-cell conditions and local impacts remain unresolved.
Dropbox agreed to buy Tigris for about $350 million in cash and unveiled DBX Agent Cloud, starting with standalone object storage. The expansion accompanies AI actions for existing file customers, while pricing and evidence of competitive performance and productivity gains remain absent from the announcement.
ElevenMusic creators can now publish directly to SoundCloud, with a three-month Pro offer for Artist Pro subscribers. The integration shortens distribution while preserving disclosure and rights requirements.
